|
Author
|
Topic: Re-Framing the Intelligent Design Issue
|
warren_bergerson
Member
Member # 262
|
posted 24. August 2002 10:33
The ‘intelligent design’ issue has traditionally been framed or structured in the form ‘Human beings, using human intelligence, produce creative designs. There exist in nature many biological systems which exhibit creative design. It is therefore reasonable to speculate that creative designs in biological systems are the result of some form of creative intelligence." This formulation of the design issue leads to the conflict with Darwinian concepts(which, it can be argued, offers a non-intelligent explanation of design.) It also creates a science/religion conflict by suggesting an explanatory role for ‘an external designer’. I am proposing here that the ‘intelligent design’ issue can be reformulated or re-framed in manner which 1)avoids, at least initially, the Darwinian and science/religion conflicts, and 2)brings the intelligent design issue into the realm of traditional scientific analysis.
I propose that as an alternative to the traditional approach, the intelligent design issue be formulated as:
1. Human beings, using human intelligence, produce adaptive designs or solutions which are sometimes creative.
2. Animals, using some process involving information processing in nervous systems, produce adaptive designs or solutions which are sometimes creative.
3. Cells, using processes which are not entirely understood, produce adaptive designs or solutions which are sometimes creative.[This refers to the short term adaptive changes in cells which may be the indirect result of longer term changes.]
4. Although the physical mechanisms involved appear to be different, all three processes are producing adaptive designs which appear to share certain logical properties including what might be labeled intelligence.
5. Systems exhibiting the complex types of adaptive or intelligent design described above do not appear to occur in nature except associated with life forms.
6. It is not unreasonable to speculate, based on the above observations, that the ability to use different physical mechanisms to produce complex, intelligent, creative designs is a defining characteristic of life forms on earth.
I believe it can be reasonably argued that the above formulation captures the essential features of the original formulation. Specifically, it captures the assertion that intelligence or intelligent design in nature is not a trait exclusive to humans. The reformulation also captures the assertion that manifestations of ‘intelligent design’ in nature are not limited to the products of human intelligence.
As I see it, there are four major advantages of the reformulation. First, the reformulation avoids direct conflict with Darwinian concepts because it addresses short term rather than long term change processes. Second, the reformulation avoids the religion/science conflict because, while it is not incompatible with the existence of a designer, it does not suggest an active role of a design in explaining the phenomena being analyzed. Third, by dealing with immediate or short term change processes (rather than events which may have occurred millions of years ago) the reformulation brings the issue into the realm of traditional scientific/laboratory analysis. Finally, by suggesting the possibility of three distinct manifestations of intelligent design, rather than two, the reformulation greatly expands the opportunities for experimental analysis.
It is a well known fact, that our ability to solve a problem depends to a very large extent on how we view or approach the problem. The ability to solve or address the intelligent design issue has, I suggest, been seriously handicapped by the way the issue has been formulated. The reformulation or re-framing of the issue proposed above avoids the conflicts which, I suggest, have prevented many scientists from seriously addressing the issue. In addition, the reformulation opens new, and potentially productive, avenues for addressing intelligent design from a scientific perspective.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 24. August 2002 15:17
All,
First of all intelligent design as proposed by Dembski/Behe does not necessarily frame the issue in terms of human beings using human intelligence. In fact the approach tries to avoid infering much about the designer. Which is in my opinion one of the major weaknesses in the ID inference since it cannot eliminate natural processes as the designer.
Secondly, I would argue that there is lack of support for the conclusion that "It is therefore reasonable to speculate that creative designs in biological systems are the result of some form of creative intelligence." In fact, it seems that processes such as RM&NS can exhibit features similar to (creative) engineering design yet they are not the result of some form of creative intelligence.
The only conflict with Darwinian concepts is that Warren seems to reach a conclusion that does not necessarily follow from the premises namely that the appearance of creative designs in biological systems is the result of some form of creative intelligence. In fact I would argue that Warren is jumping to the conclusion (creative intelligence) rather than reaching it based on the merrits of scientific analysis.
For more information on the topic of design I would like to point the reader to the very useful concepts of Doyle et al.
quote:
Nevertheless, the process of purposeless mutation and selection in our model, like biology, creates the impression of a clear direction in evolution, with results very similar to what would arise from purposeful engineering design for high yield.
From Tong Zhou*, J. M. Carlson*†, and John Doyle‡
They also have incorporated ideas of complexity and robustness and have joined 'forces' with the Santa Fe institute.
Some additional quotes:
quote:
While there are obviously fundamental differences between biology and engineering, the design and evolution processes and the resulting system-level characteristics may be much less different
than often realized. High performance lattices must have certain highly structured features such as high densities overall with barriers concentrated in high spark regions. This is largely independent of the design process, whether it be deliberate or random mutation and natural selection.
Complexity and Robustness, with John Doyle,
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 24. August 2002 18:13
Frances,
I am aware that Dembski and Behe have made extensive efforts to frame the design issue to avoid direct reference to a designer. If you read the literature, you are aware that from the perspective of academic science, the attempt has not been completely effective. My re-framing, I believe builds on their efforts in that respect.
The other major impact of my re-framing is to take Darwin and RM&NS out of the discussion of design. I think the evidence supports the observation that the rather rigid views a few indidividuals hold with respect to Darwin and RM&NS can discourage open discussion of intelligent, biological design.
Darwin and RM&NS, it will be noted, have nothing directly to say about human behavior/intelligence, adaptive learning in organisms with nervous systems, or the short term adaptive changes produced by all types of cells. The discussion/analysis of design in the way I have re-framed it is thus independent of Darwin and RM&NS.
As I have formulated the intelligent design issue, individuals starting with generalized versions of Darwinian and neo-Darwinian concepts could build models, such as variations on GA models, in attempts to model/simulate/explain occurrences of adaptive design in any of the three areas I suggested. Proposal developed from Darwinian concepts would be subjected to the same rigorous standards for precisely defining variables and operations, and for satisfying explicitly formulated fit tests as models developed from any other concepts.
Quite, frankly, it is well known from extensive testing already performed that models based on elementary Darwinian mutate-select concepts can’t pass even elementary fit tests. It is very well known that random variation and non-directed selection, can not possibly come anywhere close to simulating the type of adaptive change which is routinely observed in all three types of biological design.
Comments of the type you quote from Doyle etc. are not terribly useful in addressing the design issue. As I mentioned earlier, variance, selection, and preservation are the basic components of the Greek paradigm for teleological causation. The claim that such systems exhibit a ‘clear impression of direction’ is a restatement of a truth that has been known for a couple of thousand years, (and is probably still being taught in undergraduate courses on purpose in nature). It seems somewhat trivial to have to point out, but systems that produce short term adaptive change can not logically rely on either ‘random mutation’ or ‘natural selection’.
Let me repeat, my purpose in re-framing the design issue, is to avoid the type of ‘we already have all the answers, and we know the only way to address the problem" attitudes which are not, in general, terribly productive, and which tend to scare away many researchers who might otherwise find the issue interesting and challenging.
[ 24 August 2002, 18:21: Message edited by: warren_bergerson ]
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 25. August 2002 02:09
Doyle's work is very relevant for the design discussion since their work shows that natural 'design' can mimic intelligent design. It seems to me that Doyle et al's formulation of design has been far more succesful than ID so far or any other 'design' theory.
As long as we accept that RM&NS can lead to apparant design we have no issues here. Such acceptance however shows how difficult it may be to establish what is real versus what is apparant design.
I find it fascinating to hear that "Darwin and RM&NS, it will be noted, have nothing directly to say about human behavior/intelligence, adaptive learning in organisms with nervous systems..." when a quick search using a search engine would have established quickly that such an assertion is easily disproiven.
And may I point out once again that Warren's claims that "Quite, frankly, it is well known from extensive testing already performed that models based on elementary Darwinian mutate-select concepts can’t pass even elementary fit tests." remain once again unsupported by any supporting evidence. I realize that the moderator might object to me pointing out this reoccuring theme in Warren's arguments but I believe that it would help Warren's argument if Warren would show us some support for his claims.
I would hope that in the name of scientific research and inquiry Warren can be encouraged to support his claims. Otherwise Warren's suggestion that researchers may be discouraged by such claims as "we already have the answers" may be becoming a self fulfilling prophecy.
Let's not jump to conclusions which remain so far unsupported by factual evidence.
[ 25 August 2002, 02:19: Message edited by: Frances ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 25. August 2002 06:46
Frances,
Quote: I find it fascinating to hear that "Darwin and RM&NS, it will be noted, have nothing directly to say about human behavior/intelligence, adaptive learning in organisms with nervous systems..." when a quick search using a search engine would have established quickly that such an assertion is easily disproiven.
The three types of adaptive changes all refer to adaptive changes that can and do occur multiple times within the lifetime of an organism. Darwinian and RM&NS theory could not logically or possibly address or explain such issues since the processes and mechanisms proposed by Darwin and RM&NS are limited to mechanism that operate once per life time.
Someone could formulate an hypothesis that asserts something like -"evolutionary/adaptive changes involves a wide range of selection mechanisms in addition to natural selection, a wide range of mechanism for creating and maintaining diversity, including a wide range of non-random processes, and a wide range of processes and mechanisms for preserving adaptive information."- then such an hypothesis could be the basis for developing a theory or model that would address both short term and long term adaptive change. No one has yet explicitly formulated a modification of Darwinian theory which makes any such assertion. It is therefore logically impossible for any existing explicitly formulated form of Darwinian theory to address or explain any form of short term adaptive change.
Quote: And may I point out once again that Warren's claims that "Quite, frankly, it is well known from extensive testing already performed that models based on elementary Darwinian mutate-select concepts can’t pass even elementary fit tests." remain once again unsupported by any supporting evidence
Any intelligent, scientific discussion of any topic must assume that the participants have at least an elementary knowledge of the subject being discussed. It is well known that humans often find creative adaptive solutions to very complex problems very, very quickly( often in a single trial). It is also well known that for even slightly complex problems, pure or elementary mutate-select or GA models are extremely slow. Even the far more elaborate advanced models being evaluated today, models which are far more sophisticated than elementary mutate-select models, can not come anywhere close to meeting or fitting the speed standards of human problem solving. Your comments appear to suggest your knowledge of attempts to simulate human problem solving does not even include an understanding of such elementary facts as those outline here. As I have stated elsewhere, if you don’t understand something then ask. If you still disagree, then present some type of logic or evidence supporting your claim.
To return to the subject of this thread, re-framing the design issue, let me note, that your comments suggest that ‘re-framing’ is not the only step needed to produce a meaningful and productive dialogue on the subject of design. No matter how the issue is framed, there will be still be individuals like yourself who claim the issue must be evaluated by your peer review ‘experts’ using existing standards and using existing procedures.
It should be obvious, but probably isn’t, that there is a huge difference between ‘developing a computer program’ and ‘developing a computer program that represents a scientific model/simulation/explanation’. Developing a computer program, involves arbitrarily assigning names to variables and writing code. Developing a computer program that qualifies as a ‘scientific model’ requires logically precise definitions of variables and operations, as well as explicitly defined fit tests to determine if the model does in fact simulate the phenomena being analyzed. Scientific modeling requires the use of rigorous standards. There is not much evidence that such standards have been explicitly formulated let alone applied.
It also should be obvious, but probably isn’t, that the peer review/publish process does not work very well. Your comments suggest, that you accept a principle something like ‘If it passes peer review and is published then it is true, and if it hasn’t passed peer review and been published then it isn’t true". The flaws, the dangers, and the inadequacies of such a process in addressing complex issues of scientific modeling should, to repeat myself, be obvious. The current process is extremely slow and cumbersome. The process is easily distorted by even small amounts of bias on the part of peer reviewers. The lack of explicit peer review standards and the lack of accountability of peer reviewers insures that the system will susceptible to bias.
SUMMARY
I have proposed here for discussion a ‘re-framing of the design issue’. The re-framing was intended 1)to separate design analysis from some of the more emotional issues which, IMO, are a hindrance to productive analysis and 2)to focus on the short term aspects of design which are more readily subject to traditional forms of scientific analysis(scientific techniques are much better at analyzing phenomena that can be repeated and analyzed in the lab, than at analyzing events that may have happened once a few billion years ago). As the discussion here suggests, effective and productive analysis of intelligent, biological, design will require more than just re-framing the problem. Successful scientific analysis of design will require the development and application of rigorous scientific modeling standards. Successful scientific analysis of design will also, IMO, require a new process to replace the existing peer review/publish paradigm.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 25. August 2002 13:36
Warren comments that since "the three types of adaptive changes all refer to adaptive changes that can and do occur multiple times within the lifetime of an organism." Darwinian and RM&NS mechanisms could not address such phenomena since "mechanisms proposed by Darwin and RM&NS are limited to mechanism that operate once per life time."
This however does not follow logically. There are countless processes in an organism which take place multiple times during the lifetime of the organism and yet for those Warren does not seem to see a problem that they could in principle have arisen through Darwinian mechanisms. Why could Darwinian theory not explain the origins of systems which provide the organism(s) with increased survival?
Warren suggests that there may be more selection mechanisms than natural selection. In fact we do know of examples of artificial selection by external intelligence. One may even suggest non random processes and indeed natural selection makes for a good example of such. What other processes does Warren suggest? Evolutionary theory seems to be quite open to investigating possibilities of directed mutation and adaptive mutations but so far the findings suggest merely a process of hypermutation.
Warren responds to my request for supporting information about his claims wrt Darwinian concepts with the obvious "RM&NS is quite slow" response but that does not address his original claim that "that models based on elementary Darwinian mutate-select concepts can’t pass even elementary fit tests". I would hope that Warren will take his own advice when he states:
"If you still disagree, then present some type of logic or evidence supporting your claim."
I am thus looking forward to Warren presenting his case.
As far as how science is dealing with such issues, one may be interested to look at epigenetic inheritance for instance or The Baldwin effect.
Indeed I mentioned before Doyle et al but also would like to point out the works by Watson, Hinton
and Nowlan. Hinton and Nowlan have shown for instance that lifetime plasticity can guide evolution.
"They show how acquired traits change the shape of the reward landscape in which subsequent genetic variation takes place, and in so doing encourage the discovery of equivalent inheritable traits. This enables the seemingly Lamarkian inheritance of acquired characteristics without the direct transfer of information from the phenotype to the genotype"
Let me start to ask Warren to support another claim namely that "It also should be obvious, but probably isn’t, that the peer review/publish process does not work very well." Warren surely has guessed right here, it does not seem to be that obvious nor do I see any reason why it even should be obvious when I find little evidence that peer review process does not work very well.
Warren's strawman argument about me accepting only peer reviewed data as being 'true' further leads me to conclude that it might be beneficial for Warren to support his conclusions with references to reality.
Where did I make such claims? I merely have asked Warren to support his assertions.
In the name of science.
[ 25 August 2002, 14:54: Message edited by: Frances ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 26. August 2002 11:33
THE SPEED, STABILITY, AND COMPLEXITY PROPERTIES OF INTELLIGENT BIOLOGICAL DESIGN
Not to let this thread depart too far from the original proposal, one of the main reasons for re-framing the intelligent/biological design issue in terms of short term processes is to make it easier to apply to apply rigorous forms of scientific analysis. More specifically, I suggest, using design science techniques and the ‘short term’ approach to biological design, it is possible to develop and test scientific models and theories which "simulate, predict, and explain" at least some changes in biological design.
As a starting point, it is useful to note that one of the key elements of scientific modeling is ‘fit tests’. A mathematical model can not logically match all the feature of a complex real world process. Rather than all possible features, a scientific model attempts to match the essential features. Since, it is recognized that real world phenomena are subject to noise and inefficiency which are not applicable to ‘idealized mathematical models’, it is assumed the model must be able to outperform the real world system for the critical features. A one sided minimum performance standard for a mathematical model is called a fit test or fit standard.
Critical or essential features of biological design are defined based on survivability. A trait or performance criteria is essential to biological design if life forms could not survive without the level or standard of performance represented by the fit test. For the discussion here it is assumed that survivability and fit tests will be accepted as appropriate techniques/standards for scientific modeling of biological design.
Speed, stability, and complexity are three general properties of biological design, or of systems exhibiting the biological design property, which provide a useful basis for defining fit tests or standards. The speed standard, the requirement that an adaptive or evolutionary change must occur at a speed which permits the organism/species to survive, has been discussed under other headings. It is useful to note, there are lots of ‘directed design’ techniques which can be used to achieve very rapid adaptive change speeds.
The complexity of a system also provides a fairly obvious fit standard. A model, in order to pass a complexity fit test, must appropriately reflect the complexity of the system being simulated. Again, there has been a great deal of discussion of the mechanics of measuring complexity. It will be noted, that although biological systems are obviously extremely complex, this complexity can be addressed with relatively simple processes and mechanisms if it can be demonstrated that components of the processing are independent or parallel.
A feature of biological design which is often overlooked and/or misinterpreted by both proponents and opponents of biological design is stability. It is often mistakenly assumed that a biological design once it comes into existence, is highly stable and will continue to exist for millions and even billions of years. If biological design were stable, then design processes could in effect ‘solve it and forget it’. This stability is not the reality of biological design.
Rather than permanent or stable, biological design is highly dynamic and unstable. It is easily demonstrated that if a biological design is not modified and changed, it would in many instances quickly become mal-adaptive resulting in the death of the life form. Biological design involves processes or mechanisms operating in milliseconds to modify design in order to maintain an adaptive state( a form of design compatible with survival). As with speed and complexity, there are any number of both logical mathematical mechanisms and physical real world processes, that can operate to adjust design to maintain an adaptive state.
[As an ‘interesting’ aside, it will be noted that DNA strings or genes are an example of ‘unstable’ biological design. When cells reproduce, copying errors (mutations) produce instability. It is an interesting experiment/demonstration to start with a gene of some specified length, in a stable population, using a realistic point mutation rate and track the results over time. One can perform the demonstration either using assumed selection rates (such as survival neutral) or one can back into selection rates using what is known about the distributions and changes in distributions of alleles in a population. Using either approach, the results are dramatically different from the results supposedly produced by genetic drift theories and demonstrations. If you have a good working knowledge of this type of model you might be able to do the demonstration in your head. The suggested analysis is, at the very least, an interesting learning experience for those interested in modeling design. ]
SUMMARY
Re-framing the intelligent design issue makes it easier to demonstrate that biological design is a real phenomena that can be precisely defined and subject to rigorous scientific analysis. Furthermore, using the re-framed issue, it is possible to construct scientific models and theories which ‘simulate, predict, and explain’ at least some of the processes and mechanisms responsible for producing changes in biological design.
It may very well be true that we will ultimately find features of biological design which can not be explained by any materialistic process. It is also possible that Darwinian concepts in some form may eventually play a significant role in scientifically ‘simulating, predicting, and explaining’ intelligent or biological design. The proposed re-framing intentional avoids both ‘ultimate’ issues to concentrate on determining what we actually know and don’t know about the short term aspects of design. The evidence, IMO, is quite clear that avoiding the ultimate questions can lead to some interesting and productive analysis.
It is probably worth noting, that, again IMO, it is ‘obvious’ that scientific modeling involving precise definitions and explicit fitness tests is an appropriate method of analyzing
the different forms of biological design. It is equally as obvious that the methods currently being used in AI, in genetics, and in evolutionary biology are dramatically different from what I would characterize as scientific modeling. Models and theories are routinely presented both informally and in print as ‘explanations’ of complex phenomena without offering documentation that the models or theories meet even elementary fit standards. Models are routinely presented in AI as attempts to simulate human decision making or human problem solving, when such models are clearly diverging from rather than converging on what is known about human decision making performance or human problem solving performance.
There are serious technical issues which need to be analyzed in depth relating to what can and can not be accomplished modeling biological design. Such serious discussion will not occur starting with the premise AI or evolutionary biology or genetics already has all the answers. Neither will the issues be resolved by arguing the answers already exist in the published peer reviewed literature.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 26. August 2002 12:13
I applaud Warren for his efforts regarding 'design' but there still remain some claims that I believe could benefit from additional information, references, worked out examples, simulations etc.
For instance, Warren could explain how his claims that "Rather than permanent or stable, biological design is highly dynamic and unstable. It is easily demonstrated that if a biological design is not modified and changed, it would in many instances quickly become mal-adaptive resulting in the death of the life form. " can be reconciled with the so often found robustness of 'biological design'? In fact I am not so sure that biological design in many instances would become mal-adaptive. It might be helpful to introduce some relevant time scales for instance.
As far as Warren's ideas of complexity, speed and stability are concerned, he might want to check out the work by Doyle for instance or the work by Lenski/Adami
As far as speed is concerned, I have pointed out several examples which show how Darwinian theory deals with variations on shorter time scale than the average life cycle.
Finally I would be very interested in seeing Warren provide some worked out examples for a point mutation in a stable population that show that the "results are dramatically different from the results supposedly produced by genetic drift theories and demonstrations". Did genetic drift theories predict something that such simulations show to be erroneous? If that is the case then some assumption must surely have been violated. Theoretical population biology surely could not be that wrong?
[ 26 August 2002, 12:33: Message edited by: Frances ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 26. August 2002 13:20
Frances,
I don’t entirely understand where your comments from your earlier posting are heading.
I proposed that the intelligent/biological design issue be framed initially in terms of short term adaptive changes, in part to avoid the conflict with the somewhat inflexible views of some supporters of certain evolutionary theories. Clearly, some individuals might find Darwinian concepts useful in developing models of short term -within lifetime adaptive changes( such as learning). I was not aware, however, that any such model or theory has been explicitly formulated and offered for testing and review.
Also with respect to ‘can’t pass elementary fit tests’, I am not sure if you are saying 1)you don’t understand the math, 2)you have a counter-demonstration that contradicts the conclusion or 3)you don’t accept the type of fit tests proposed. If you clarify your position, maybe we can discuss it.
I am not sure if the adequacy of the ‘peer review/publish’ process is an appropriate subject for discussion here. However, let me note that my comments relate to specific rather than general issues. First, peer review/publish does not provide an effective method of evaluating the impact of subsequent discoveries on past conclusions. Second, peer review provides no mechanisms or standards for controlling post publication use and misuse of published material. Third, peer review does not have explicitly stated, reviewable standards for acceptance or rejections of models and theories. Fourth, peer review does not have rigorous, explicitly stated reviewable standards for general assertions regarding what is and what is not explained by certain theories. I could go on.
Stability of design depends in part on how it is defined. You can attempt to define biological design in a manner which suggests stability, but then you can’t explain how it changes. As simple examples of dynamics and instability as defined here, learning is an obvious example of redesign. Failure to learn any number of things can lead to the death of an animal. If the ‘design’ of individual cells did not change or readapt over time, a multi-cellular organisms would fail to develop and would die. Both individual cells and complex organisms are highly adaptive. They are capable of rather dramatic changes and failure of any one of many such adaptive processes would result in death.
Quote: Theoretical population biology surely could not be that wrong?
Do the math, which isn’t particularly complex. Unless you assume very severe selective forces, a gene in a population will diverge very rapidly into many different alleles. If on the other hand, you assume severe selection(elimination of essentially all divergent forms), you get essentially no diversity. There is, as your question suggests, there is a ‘second’ ‘interesting’ ‘human behavior’ aspects to this issue. Get someone to do/confirm the math, and I will be glad to discuss the second issue.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 27. August 2002 02:15
Warren has raised an interesting point when claiming that an animal has to be quite adaptable at short time scales. But to suggest that they are capable off dramatic changes seems to require some additional information. First of all are these changes in the phenotype or the genotype. As I have shown to Warren, evolutionary theory has quite a good handle on some of these concepts, I mentioned I believe the Baldwin effect.
I would suggest that Warren takes a look at this interesting concept which seems to be quite well reconcilable with Darwinian theory.
I hope that me making Warren aware of the vaste amount of research that has researched learning and within lifetime changes. I am glad that I have been able to contribute to making Warren aware of these concepts.
Once one starts digging beyond the depths of strawmen one will find that Darwinian theory is actually quite capable of explaining many of the observations.
It's no wonder that the theory has withstood so well through time.
As far as doing the math is concerned, I wonder if Warren is familiar with 'genetic drift' for instance. Theoretical biology is quite familiar with modeling but I would like for Warren to explain his comment that
'Finally I would be very interested in seeing Warren provide some worked out examples for a point mutation in a stable population that show that the "results are dramatically different from the results supposedly produced by genetic drift theories and demonstrations". "
What is Warren trying to say here? That genetic drift can play a role when population sizes allow for this? Perhaps Warren may want to explain the relevance of his comments?
Warren started off with questioning if I did not understood the math. My response to his request is simple: What math? Warren has certainly not shown any relevant math. I have asked Warren before to support his claims that 'that models based on elementary Darwinian mutate-select concepts can’t pass even elementary fit tests". '
Surely Warren does not think that he can explain the absence of such supporting evidence by suggesting that *I* do not understand 'the math'
?
As far as stability is concerned, Warren is wrong to suggest that one cannot formulate biological design in a manner which explains stasis (stability) as well as change. In fact Doyle et al have done exactly this, robustness of systems yet still susceptible to external variations. In fact it's fascinating to see how PunkEek and stasis can be understood by the same theoretical approach.
Is learning an example of redesign? Redesign of what? Phenotype, Genotype? What is being redesigned? Is the term design even useful in such context? As far as multicellularity is concerned it seems that adaptation to predation might have been an explanation for it. In fact this seems quite understandable in Darwinian terms.
Funny how variation and selection seem to be such powerful concepts that we now see them be incorporated in 'intelligent design' approaches. We have a lot to learn from nature.
[ 27 August 2002, 02:31: Message edited by: Frances ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 27. August 2002 10:28
As I stated earlier, there is a simple mathematical experiment/demonstration that you can do at home that shows that mutate-select does not operate as suggested by existing theory. The following are a few details to make it easier to get started.
To begin, this demonstration shows that you can have 1)a documented/observed pattern of change(genetic drift ‘the fact’ or ‘chance changes in the distribution of alleles’ does occur,) 2)mathematical models which simulate random change(such models are easily created and published) and 3) a theory or model which does not and can not explain, model or simulate the observed phenomena. It is important to understand that you can create a genetic drift type pattern with a mutate-select type model, but you can not explain, model, or simulate actual occurrences of genetic drift with a realistic random mutation and natural selection model.
The demonstrations of genetic drift the ‘theory’ which routinely pass ‘peer review’ are based on unrealistic, or ‘artificial’ assumptions regarding mutation and selection rates. In most fields of endeavor, the use of such artificial assumptions would not be permitted. But you don’t need to take my word for it. You can easily work through the math in the following demonstrations.
GENETIC DRIFT DEMO
As a starting point, assume a stable population of N(10,000?), a gene or string of DNA with M(20,000?) units or pairs, and a point mutation rate of R (1 per 10 million cell divisions?), a cell division rate of S (once per month?), and an average life time of T (40 years?). The specific values of N, M, R, S, and T are not critical as long as they are reasonably consistent with what might occur in nature.
The first experiment, assume the trial starts with the entire population having two closely related forms or alleles A and B of the gene with the population roughly evenly divided between the two forms. Assume that all forms of the gene are ‘selection neutral’. What distribution of alleles would be expect to find after a 100 or 1000 generations?
Genetic drift theory suggests that as a result of random or chance drift, the frequency of A might have increased, decreased or stayed the same with the opposite impact on B.
Our simulation, by contrast, suggests that if the gene is truly ‘selection neutral’, the distribution of both A and B would reduce dramatically and be largely replaced by a ‘random’ mix of the roughly 3*M one mutation variations of A and B. I would suggest this phenomena be called ‘destructive divergence’ rather than genetic drift. I would suggest that this might be what happens to the genes controlling the eyes of fish when they live in caves where there is no light.
The second part of the experiment accepts that the frequency of alleles A and B are stable or drift over the period of the study. The study then ‘backs into’ the selection factors or forces needed to explain the observed drift. The results obtained clearly show that all of the 3*M ‘common’ mutations (other than mutations changing A to B or B to A) are immediately eliminated. If we assume that this selection is a result of Darwinian natural selection, then all cells with one of the 3*M mutations other than A or B will die and/or fail to reproduce.
This does not, however, appear to be possible. If we look at the size of the genome, and at the average number of point mutations likely to occur in each cell division, and if the ratio of lethal to non-lethal mutations is something like 30,000 to 1, then very few cell divisions would ever result in the production of a living cell. If, in effect, the theory of genetic drift is correct, then the organism will not survive.
The obvious alternative, is some simple type error elimination process. But if such mechanism exist to the extent suggested by this demonstration, then the process also severely limits the number of possible mutations. If, as in the example here, there are only two possible mutations, then it would seem highly inappropriate to describe the mutation process as random.
WHAT DOES IT MEAN?
This demonstration shows the rather obvious fact that there is a very big difference between ‘drift in a very stable artificial environment’ and ‘apparent drift and stability in a an unstable, dynamic environment’. To use a real world analogy, there is a very large difference between a ‘piece of cloth hanging on a clothes line’ and ‘a man balancing on a tight rope’. One is an example of a simple static system. The other is an example of dynamic adaptive system that is continually readapting in order to maintain the impression of stability or equilibrium.
Genetic drift, to be kind, is a ‘flawed’ theory which gained acceptance because no one applied an elementary ‘stability’ fit test. How could such an ‘obvious’ error continue to pass the peer review process? The answer, of course, is that the process is seriously and fundamentally flawed. Not only have individuals performing peer review failed to apply an obvious and critical fit test, but there is no indication that they are even aware such a test exists or applies.
Darwin and company asked the ‘interesting’ question "How and why do species change?" The equally interesting and more overlooked question is "How and why do species remain the same in an environment that should produce divergence?" . The question is not only, how does the man or woman walk across the tightrope, but how does he or she keep from falling off. Darwin and company developed theories that might possibly work in a very static, very stable environment. Darwinian and neo-Darwinian theories do not work in a real dynamic, unstable environment and they do not pass elementary stability fitness tests.
RELEVANCE TO THE RE-FRAMING
Biological systems have the ability to maintain adaptive states despite high levels of instability in the environments in which these systems exist. When viewed from a long term perspective, the ability to maintain adaptive states or adaptive equilibrium gives the appearance of ‘a rock sitting on the ground’. Only by looking at the short term dynamics of design does it become obvious that ‘staying alive’ is a whole lot more complex than ‘a rock sitting on the ground’. Looking only at the long term aspects of biological design provides a very distorted and incomplete view of the processes involved.
The above genetic drift illustration also provides a clear demonstration of the 1)scientific standards issue , and 2)the scientific process issue. It is, IMO, fairly obvious that you can not develop scientific models or theories which can ‘simulate, predict, and explain’ changes in design unless you establish and apply fit tests including stability fit tests. It is also fairly obvious that such fit tests have not been developed and are not being applied in either evolutionary biology or genetics.
There are two options for addressing these issues. First, one can go to the scientists and peer reviewers in the various fields of biology and convince of the need to develop and apply fit standards in evaluating the soundness or validity of a scientific model or theory. Second, one can establish a new science which develops and applies fit standards and demonstrate the new science produces useful results. Ultimately, both options need to be addressed since different sciences must be in agreement on the acceptance or rejection of a scientific theory or model. My proposal for re-framing the design issue was an attempt to delay the conflict with evolutionary biology and genetics on the question of fit standards, and to start by developing a new design science which used the more rigorous standards.
Frances is suggesting that in analyzing short term changes in design we must follow the existing standards and processes from evolutionary biology and genetics. I do not advocate such an approach because, as demonstrated above, it leads almost immediately to questioning the adequacy and soundness of existing standards and procedures. Although I am not unwilling to undertake such a debate, I think it would be more productive to discuss/address the many interesting issues raised by attempting to analyze the short term features of biological design.
FRANCES,
You raise the ‘if you looked at the literature you would understand’ argument. As I have said on numerous occasions, no one doubts that biologists are excellent observers and that biologists from a variety of disciplines have accumulated a very impressive body of knowledge. But the issue here is ‘the formulation and testing of scientific models and theories’ not the accumulation of facts. Biologists routinely identify and assign names to phenomena. This, it should be obvious, is not the same as formulating a testable scientific model and theory to simulate and explain the phenomena. Nor is it the same as demonstrating the observed phenomena can be modeled and explained in terms of Darwinian theory. The ‘if you looked…’ argument, in this forum, is more like a ‘I can’t come up with or find a good argument to support my position, but I am sure if you read a few million published articles you could find a good argument to support my position’.
On a somewhat more substantive subject, you raise questions such as "What is being redesigned?" and "Is learning an example of redesign?" Another of the key issues in scientific modeling is precisely defining terms and concepts. One of the original strengths of neo-Darwinian genetics was to propose/offer an explicit and precise definition of ‘point in time design’ based on genes and DNA. Early efforts also attempted to provide precise definitions of selection and mutation. Unfortunately, the early efforts at precise definition appear to have been abandoned to the point that people now use terms like mutate and select to refer to whatever is convenient at the moment.
In order to develop scientific models and theories of short term changes in design, one must develop explicit and logically precise definitions of ‘point in time design’. I am suggesting that there are (at least) three types of biological design, cellular, nervous system and human. In order to develop scientific models of changes in each type of design, it is first necessary to develop explicit and logically precise definitions of ‘the status of each type of design at a point in time’ and an explicit definition/model of what changes when design changes. Although I have developed such definitions and models, they have not been discussed here. Defining and modeling point in time design is a fairly technical subject and I would be glad to discuss it with anyone with a serious interest in the technical aspects of the subject. Frances, as your questions suggest, precisely defining terms and processes is a critical feature of modeling.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 27. August 2002 23:10
I am confused, Warren seems to label "GENETIC DRIFT DEMO" and then continues to propose a population of 10000 organisms. Yet it should be clear that for genetic drift to play a role, the initial population needs to be small. Furthermore his 'demo' seems to be somewhat unclear about what he is trying to say. Warren says that the distribution of A and B would reduce dramatically to a mix of 3M mutation variations? I would certainly love to see the underlying logic to support this assertion? How did Warren reach the 3M number and why would there be 3M variations? Is Warren really saying that there are 60,000 variations of A and B in a population of 10,000?
Something just does not seem to add up here.
Even more interesting seems to be Warren's assertion that all cells with mutations other than A or B would die or fail to reproduce. Perhaps Warren could explain to us how he reaches this conclusion? From Warren's examples I am not sure if we are discussing the same biology (and mathematics) here as used in theoretical population genetics for instance.
Even more interesting is Warren's assertion that if random drift were correct, most organisms would not survive.
Given the controversial nature of Warren's claims I would like to encourage Warren to either 1) show the simulation 2) provide for the mathematical derivations for the model.
Warren then concludes that there is a difference between drift in a stable environment and drift in a dynamic environment. An interesting assertion but I fail to see how this has any relevance to our discussion.
Before Warren suggests that genetic drift is flawed, I encourage him to actually show that this is the case.
Once Warren proposes quantifiable calculations can we determine if there are some/any merits to his claims.
I will be awaiting Warren's quantifiable contribution with much anticipation.
So far though I have to admit that I am somewhat disappointed. Warren has made several assertions about RM&NS which so far remain unsupported and I fail to see how his present proposal will provide for a better description of the known data.
Perhaps with Warren's help we can resolve these issues.
[ 27 August 2002, 23:25: Message edited by: Frances ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 28. August 2002 13:17
COMPARING OF NEW AND EXISTING APPROACHES
Even though the point may be ‘obvious’, it is worth noting that there are both ‘technical’ and ‘non-technical’ issues that go into a ‘decision’ to 1)reject an existing, established approach and 2)pursue a new untried approach. The (re-frame design/design science) approach I advocate is new approach intended as an alternative to existing approaches to ID, and more fundamentally as an alternative to the existing approaches to the analysis of evolutionary/adaptive change.
On a purely technical basis, a strong argument can be made for seriously considering the new approach. One of the ‘goals’ of scientific analysis is the formulation of testable, predictive, ‘hard science’ models and theories which are capable of ‘simulating, predicting, explaining’ the phenomena being analyzed. Existing approaches have clearly not succeeded in producing testable, predictive, hard science models and theories of evolutionary/adaptive change. The new re-framing/design science approach, if any one is interested in looking, offers specific solutions to a number of the technical problems associated with formulating predictive models including 1)fit tests and 2)techniques for defining/measuring the information/design property. On a purely technical basis, evolutionary biology, and in fact all of the life sciences, should be very open to any new approach which offers a reasonable possibility of producing hard science predictive theories.
The non-technical basis for accepting or rejecting a new approach tend to be more ‘personal, parochial, and pragmatic’. An individual scientist is more likely to reject or ignore a new approach if the new approach 1)requires skills the individual doesn’t have or can’t readily acquire, 2)produces conclusions contrary to the views previously expressed by the individual, 3)has the potential to negatively impact the prestige of the individual, 4)has the potential to produce a negative financial impact on the individual,… etc. There is ample evidence from any number of fields that these non-technical, personal interest factors can seriously distort the views, opinions, and/or decisions of essentially any group of individuals. (Seriously distorted advice from ‘investment experts’ provides one of the more dramatic recent examples of this phenomena.)
This is not the place to discussion the ‘non-technical’ aspects of acceptance/rejection of a proposal and it is not my intent to initiate such a discussion. I simply wish to point out that the approach being proposed here involves the use of modeling techniques not currently in common use and new technical/mathematical fit tests to determine the validity of models. The use of these ‘new’ techniques and tests produces results which are contrary to published and peer reviewed opinions. The proposals offered, if they are followed to their logical conclusion, suggest the need for changes in the institutional procedures used to present and review materials. Given the nature of the proposals being made, it seems likely that some individuals will reject or ignore the proposal on ‘non-technical’ grounds.
Frances,
In attempting to discuss a technical issue, it is not uncommon to get irrelevant, non-substantive questions or comments. The points you raise with respect to the genetic drift demonstration all appear to be in this irrelevant and non-substantive category. When this occurs, and in any technical discussion it is not an uncommon occurrence, it is reasonable to ask are the irrelevant questions and comments, 1)due to ambiguous features in the initial presentation, 2)the inability of the individual raising the questions to understand the issue, or 3)an intentional attempt to disrupt the discussion. Although I am sure many individuals might have difficulty following the logic of the argument presented, your questions and comments clearly suggest you are either unable or unwilling to seriously discuss the technical aspects of the demonstration.
IP: Logged
|
|
Frances
Member
Member # 169
|
posted 28. August 2002 13:59
I am amazed and somewhat disappointed that Warren, when asked to support his claims about for instance genetic drift resorts to claims that this is irrelevant or non-substantative.
From my previous discussions with Warren as wekll as from reading Warren's comments on other boards I have come to the conclusion that Warren often seems to be making claims which when asked to support them, remain unsupported. This does not mean that the claims themselves are unsupportable, but it shows that Warren needs to provide some evidentiary support for his claims. I have tried to direct the discussion back to supporting the evidence but Warren's response suggests that he is not interested in pursuing such a track.
Instead Warren tries to imply that I am here to disrupt. If disruption means to ask for supporting evidence then I plead guilty but then scientific inquiry would be equally guilty of 'disruption'. Perhaps, disruption in this sense is actually helpful in showing the problems with certain claims.
I do agree though with Warren that irrelevant comments or questions may be evidence of ambiguity, inability to understand the issues. Am I the only one to see the irony of Warren's comments here?
:-)
But to accuse me of not willing to seriously discuss the technical aspects ignores the facts that I have been the only one so far willing to do so.
If Warren is willing to make his case in a scientific format then I will be more than happy to discuss with him his ideas.
Looking forward to such a discussion.
====Added====
I think it is important to realize that when proposing a new or alternative idea that one at a minimum shows that the new idea explains the existing data better than the old idea. Warren seems to reject RM&NS for reasons that remain sofar unclear to me and yet proposes an 'alternative' which in many aspects does not seem to be adding much, certainly it has not been shown that the new idea is a better idea. By itself that is no reason NOT to pursue the new idea and see where it leads though but I would like to understand why Warren sees the need for a new idea when the old idea has yet to be shown to be failing?
[ 28 August 2002, 14:44: Message edited by: Frances ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 29. August 2002 09:57
Frances,
A couple of days ago I posted mathematical demonstrations that show:
1. The theory of genetic drift, given what is currently known of genetic mechanism, is a mathematical, logical impossibility. The conditions which in theory would produce drift, would instead produce divergence.
2. The RM&NS theory, given current knowledge of genetics, is a mathematical, logical impossibility. All existing life forms would die if RM&NS was true.
3. Darwin’s theory of natural selection, given current knowledge of genetics, is a mathematical logical impossibility. Life and evolutionary change require the existence of selection mechanisms other than those included under the heading of natural selection.
You continues to make claims that the mathematical demonstrations are not complete or valid. I checked with acquaintances who are mathematicians (actuaries) and they confirmed that the mathematical demonstrations presented are clear and the mathematical conclusions obvious. Your factual claims regarding the soundness of the mathematical demonstrations presented are wrong, and without logical, mathematical or scientific foundation. Unless you or someone else raises new and substantive issues, there is little more that can be done or needs to be done in support the above conclusions. I ask, as a matter of courtesy, that you quit making the unsupported accusations.
[Moderator: It appears, in the absence of contrary evidence or arguments, that the claims made above can be supported based on merit. It is also obvious that these conclusions are contrary to views and beliefs almost universally accepted in biology. Although it is contrary to what is normally considered appropriate, I would like to propose (or if you prefer speculate on) an explanation of why Frances and essentially the entire scientific community can support, or appear to support, ‘scientific theories’ which directly contradict mathematical logic and known facts.]
The following, IMO, are some of the major reasons that Darwinian and neo-Darwinian theories are widely accepted despite the existence of contradictory evidence:
1. KNOWLEDGE OF BASICS- Mathematics, mathematical modeling, theory construction, theory testing are complex technical skills requiring relatively rare human skills and extensive development of those skills. While many individuals have some skill and some experience with mathematics and scientific modeling, very few individuals, even very few mathematicians or scientists, actually have the skills and experience to formulate and evaluate scientific/mathematical models and theories. Although the demonstrations I presented involve ‘relatively elementary math’ and ‘relatively elementary issues of theory construction’, unless an individual has direct experience/exposure to the issues involved, understanding them will require an effort.
2. RELIANCE ON AUTHORITY/CONFORMITY- In most instances, our views and beliefs, and this includes the acceptance of ‘scientific truths’, rely heavily on the analysis and conclusions produced by experts. This reliance on authority is so strong that in many, easily documented instances, individuals accept the authoritative view, even when it contradicts their own experience and analysis. Reliance on authority also leads individuals to believe in or assume the existence of rigorous ‘scientific evidence’ when no such evidence exists.
The above two factors explain, IMO, what can be called the ‘popular views and beliefs’ surrounding Darwinian and neo-Darwinian theories. [ Reliance on authority and/or "views and beliefs based on ‘unquestioning’ reliance on authority" is far more widespread than is generally recognized. Most individuals, including most scientists, base most of their views and beliefs on ‘reliance on authority’ rather than on actually performing or even reviewing the evidence. ]
While ‘knowledge of basics’ and ‘reliance on authority’ might explain the positions of many individuals, including the views of many scientists, they clearly do not explain the views of knowledgeable scientists actively involved in analyzing evolutionary theory. To explain the views of ‘scientific experts’ we need to look to the following three factors:
3. LITERAL AND MATERIAL TRUTH- No serious scientist would, as far as I know, seriously suggest that Darwinian or neo-Darwinian theories were ‘literally’ true. Essentially everyone knowledgeable on the subject realizes that evolutionary change processes are more complex than and involve processes and mechanisms other than the mechanisms described/defined by existing theories. The demonstrations I provided show in part that existing theories are not literally true. While such demonstrations may be interesting they are neither surprising nor particularly significant. The ‘stability fit test’ on which the demonstrations are based is of more general significance because it suggests that there are material flaws in existing theories.
4. EXISTENCE OF ALTERNATIVE THEORIES- The rule of etiquette in science is not to reject one theory unless a valid alternative can be offered. While it is relatively easy to disprove and/or identify shortcomings in existing theories, it is has proved very difficult/challenging to formulate an alternative model or theory which both fits the large body of existing knowledge and produces meaningful testable predictions. This problem is not one unique to evolutionary biology, but one applicable to all attempts to formulate theories of complex real world ‘paradigms’. By some strange perversion of logic, this ‘no one knows how to formulate a valid alternative’ has gotten interpreted to mean ‘the existing theories are valid’. The design science I am promoting appears to offer some technical solutions to the problem of formulating scientific models and theories of paradigms or ‘complex/diverse sets of causal relationships’. The popular, ‘we already have an answer’ viewpoint makes it difficult to even discuss the problem of formulating alternative theories.
5. BUSINESS/FINACIAL BIAS - On top of the complex technical issues associated with evaluating evolutionary theory, are the day to day business and financial concerns. Scientists need to get published, they need funding for their research, they ‘need’ to be accepted and respected by their peers. As has been shown in area after area, these concerns play a significant role in how views are expressed and slanted. At the very least, such concerns lead individuals to express views in forms that can be interpreted as compatible with the interests of the ‘business of science’. Although, most people believe ‘it would never happen to us’, it is very common for non-technical, personal-interest concerns to eventually influence and override technical, scientific issues.
SUMMARY
As support for the argument that ‘re-framing the design issue’ could lead to productive analysis, I presented the concept of the ‘stability fit test’. In support of the value of the stability fit test I presented a mathematical demonstration that produced the conclusions stated at the beginning of this post. The soundness, completeness, and validity of the demonstrations have been questioned, but those claims, based on the evidence of individuals with relevant knowledge, do not appear to be substantive. I have attempted above to provide an explanation of ‘how’ a seemingly simple demonstration can produce results which contradict the generally accepted views, or what appear on the surface to be the generally accepted views of the academic/scientific community.
IP: Logged
|
|
|
Printer-friendly view of this topic